Obstacle detection method and system for vehicles traveling in underground coal mine tunnels

Abstract

The present invention provides a method and system for detecting obstacles in front of a vehicle traveling in an underground coal mine lane, and relates to the technical field of underground mine image recognition, including collecting vehicle vibration parameters, reference images and comparison images, and obstacle recognition images, and performing correlation analysis, generating a comprehensive clarity evaluation index for image recognition of vehicles traveling in the mine, and outputting relevant accuracy levels; the present invention is the first to evaluate and determine three factors that cause the obstacle recognition images taken by a camera to be relatively blurred and affect the accuracy of obstacle recognition in the process of underground mine vehicle driving and identifying obstacles, and by collecting vehicle vibration parameters, reference images and comparison images, and obstacle recognition images, performing correlation analysis, generating a comprehensive clarity evaluation index for image recognition of vehicles traveling in the mine that reflects the accuracy of obstacle recognition in the underground coal mine lane, and outputting relevant accuracy levels to assist in determination, thereby improving the accuracy of image recognition of obstacles in front of vehicles traveling in underground coal mine lanes.

Description

Technical Field

[0001] The present invention relates to the technical field of underground mine image recognition, and in particular to a method and system for detecting obstacles in front of a vehicle traveling in an underground coal mine tunnel. Background Art

[0002] When detecting obstacles in front of vehicles traveling in underground coal mine tunnels, images are collected and obstacles are identified through on-board cameras. Image recognition in underground coal mines requires the use of computer vision, image processing, and deep learning technologies, combined with professional knowledge and experience in the coal mining industry. By collecting a large amount of image data, annotating it, and training a deep learning model, it is possible to accurately identify and analyze underground coal mine scenes, thereby identifying and analyzing obstacles for traveling vehicles. During the identification process, images need to be collected through on-board cameras.

[0003] While moving, the vehicle vibrates irregularly due to the complexity of the tunnel road and its own factors, which causes the on-board camera to shake irregularly, thus affecting image acquisition; at the same time, the dust concentration underground is relatively high, which affects the visibility in the mine, thereby affecting the clarity of the obstacle recognition image collected by the camera; in addition, due to the deviation of the camera's focus on obstacle recognition, the obstacle recognition image taken by the camera is relatively blurred, affecting the accuracy of obstacle recognition; the above three factors will lead to a decrease in the accuracy of obstacle recognition in front of vehicles traveling in underground mine tunnels.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for detecting obstacles in front of a vehicle traveling in an underground coal mine tunnel, so as to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for detecting obstacles in front of a vehicle traveling in a coal mine tunnel is applicable to collecting images and identifying obstacles through a vehicle-mounted camera, and the specific steps include:

[0008] Step 1: collecting vehicle vibration parameters of a vehicle traveling in a mine, wherein the vehicle vibration parameters include the vehicle vertical acceleration and lateral translation acceleration;

[0009] Step 2: Perform correlation analysis on the vehicle vibration parameters to generate the vehicle vibration speed variation coefficient;

[0010] Step 3: collecting a reference image of the traveling vehicle as a baseline reference, collecting a comparison image of the traveling vehicle traveling in a mine as a comparison reference, and collecting pixel values ​​of reference objects in the reference image and the comparison image;

[0011] Step 4: Perform correlation analysis on the grayscale values ​​of the reference objects in the reference image and the comparison image to generate a reference environmental impact coefficient;

[0012] Step 5: Collect obstacle recognition images taken by the camera of the vehicle traveling in the mine;

[0013] Step 6: Perform correlation analysis on the obstacle recognition image to generate pixel mean, pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation. Perform correlation analysis on the pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation to generate a no-reference image clarity evaluation coefficient.

[0014] Step 7: Conduct correlation analysis on the clarity assessment coefficient of the non-reference image, the reference environmental impact coefficient, and the vehicle vibration speed variation coefficient to generate a comprehensive clarity assessment index for mine vehicle image recognition. Compare the comprehensive clarity assessment index for mine vehicle image recognition with the impact threshold and output the accuracy level of obstacle image recognition in front of vehicles traveling in coal mine tunnels.

[0015] Furthermore, in step 1, the vertical acceleration A of the vehicle is collected by a three-axis acceleration sensor installed on the moving vehicle. Y and lateral translation acceleration A X , A X A is the acceleration value of the vehicle perpendicular to the direction of travel of the vehicle and parallel to the horizontal plane, Y It is the vehicle acceleration value in the vertical direction of the moving vehicle.

[0016] Furthermore, the vehicle vertical acceleration A Y and lateral translation acceleration A X Correlation analysis is performed to generate the vehicle vibration speed variation coefficient CDX, based on the formula: Wherein, ε is the vehicle acceleration jitter factor, and the value range of ε is [0.35, 3.88]. The vehicle jitter speed variation coefficient reflects the severity of the lateral and longitudinal comprehensive jitter of the moving vehicle.

[0017] Furthermore, in step 3, a camera and a reference object are set outside the moving vehicle, and the distance between the reference object and the camera is greater than 50 cm and less than 150 cm. When the mass concentration of PM2.5 is less than 50 μg·m 3When the relative humidity of the air is less than 60%, the camera is used to shoot a reference object to generate a reference image. The reference object in the reference image has M rows and N columns of pixels, and the grayscale value of the pixel in the m0th row and n0th column of the reference object in the reference image is recorded as I(m0,n0); during obstacle image recognition and detection, the camera is used to shoot a reference object to generate a comparison image. The reference object in the comparison image also has M rows and N columns of pixels, and the grayscale value of the pixel in the m0th row and n0th column of the comparison image is recorded as G(m0,n0).

[0018] Furthermore, a correlation analysis is performed on the grayscale values ​​of the reference objects in the reference image and the comparison image to generate the reference environmental impact coefficient HYX, based on the following formula: The reference environmental impact coefficient reflects the degree of influence of the underground mine environment on the shooting clarity.

[0019] Furthermore, in step 5, the obstacle recognition image has X rows and Y columns of pixels, and the brightness of the pixel in the i-th row and j-th column is represented by f(i, j).

[0020] Furthermore, the obstacle recognition image is subjected to correlation analysis to generate pixel mean L, pixel grayscale distribution discrete value H, horizontal pixel variation T X , vertical pixel change degree T Y , the formula based on is:

[0021]

[0022] The pixel mean L reflects the average brightness of the obstacle recognition image, the pixel grayscale distribution discrete value H reflects the brightness discreteness of the pixel points of the obstacle recognition image, and the horizontal pixel variation T X It reflects the brightness change degree of pixels separated by two pixels in the horizontal direction of the obstacle recognition image. The vertical pixel change degree T Y It reflects the degree of brightness change of pixels separated by two pixels in the vertical direction of the obstacle recognition image.

[0023] Furthermore, correlation analysis is performed on the discrete values ​​of pixel grayscale distribution, horizontal pixel variation, and vertical pixel variation to generate a reference-free image clarity evaluation coefficient TQP, based on the following formula: Wherein, α is the brightness discrete influencing factor of the obstacle recognition image, β is the gradient discrete influencing factor of the obstacle recognition image, and β>α≥6.22. The no-reference image clarity evaluation coefficient TQP reflects the clarity of the obstacle recognition image.

[0024] Furthermore, the correlation analysis of the non-reference image clarity evaluation coefficient, the reference environment impact coefficient, and the vehicle vibration speed change coefficient is performed to generate the mine vehicle image recognition comprehensive clarity evaluation index KSQ, based on the formula: The comprehensive clarity evaluation index of mine vehicle image recognition reflects the accuracy of obstacle image recognition in front of vehicles traveling in underground coal mine tunnels.

[0025] The present invention also provides a system for detecting obstacles in front of vehicles traveling in underground mine lanes, which is used to execute the above-mentioned method for detecting obstacles in front of vehicles traveling in underground mine lanes, and comprises:

[0026] An acceleration acquisition module, which is used to collect vehicle vibration parameters of vehicles traveling in the mine;

[0027] A first analysis module, the first analysis module is used to perform correlation analysis on vehicle vibration parameters to generate a vehicle vibration speed variation coefficient;

[0028] An environment change model establishment module, the environment change model establishment module is used to set a camera and a reference object outside the moving vehicle;

[0029] An environment change acquisition module, the environment change acquisition module is used to acquire a reference image of a moving vehicle as a baseline reference, and to acquire a comparison image of the moving vehicle when it is traveling in a mine as a comparison reference;

[0030] A second analysis module, the second analysis module is used to perform correlation analysis on the grayscale values ​​of the reference image and the comparison image to generate a reference environmental impact coefficient;

[0031] An image recognition acquisition module, which is used to acquire obstacle recognition images taken when the vehicle is in motion;

[0032] a third analysis module, the third analysis module being used to perform correlation analysis on the obstacle recognition image, generate pixel mean, pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation, and perform correlation analysis on the pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation to generate a no-reference image clarity assessment coefficient;

[0033] A comprehensive analysis module, which is used to perform correlation analysis on the vehicle jitter speed variation coefficient, the reference environment impact coefficient, and the non-reference image clarity evaluation coefficient output by the first analysis module, the second analysis module, and the third analysis module, and generate a comprehensive clarity evaluation index for mine vehicle image recognition;

[0034] A comparison module, wherein the comparison module is used to compare the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine with the impact threshold, and output the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine; when the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine is greater than the impact threshold, the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine is output as level one; when the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine is less than the impact threshold, the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine is output as level two.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention is the first to evaluate and determine the three factors that cause the obstacle recognition images taken by the camera to be blurred and affect the accuracy of obstacle recognition during the process of vehicle driving and identifying obstacles in mines. The present invention evaluates and determines the three factors for the irregular vibration of the driving vehicle, which causes the on-board camera to shake irregularly, and the dust concentration in the mine to be high, which affects the visibility in the mine. The present invention also determines the deviation in the focus of the camera on obstacle recognition. The present invention collects vehicle vibration parameters, reference images, comparison images, and obstacle recognition images for correlation analysis, generates a comprehensive clarity evaluation index for image recognition of vehicles driving in mines, which reflects the accuracy of image recognition of obstacles in mines, and outputs relevant accuracy levels to assist in judgment, thereby improving the accuracy of image recognition of obstacles in front of vehicles driving in underground mine lanes. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0038] Figure 2 It is a schematic diagram of the overall system module of the present invention. DETAILED DESCRIPTION

[0039] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0040] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0041] Example:

[0042] See also Figure 1 , the present invention provides a technical solution:

[0043] A method for detecting obstacles in front of vehicles traveling in underground coal mine lanes is suitable for collecting images and identifying obstacles through vehicle-mounted cameras. Image recognition in underground coal mines requires the use of computer vision, image processing, and deep learning technologies, combined with professional knowledge and experience in the coal mining industry. By collecting a large amount of image data, annotating it, and training a deep learning model, it is possible to accurately identify and analyze underground coal mine scenes, thereby identifying and analyzing obstacles for traveling vehicles. During the recognition process, it is necessary to collect images through the on-board camera. However, when the vehicle is moving, the complex road and its own factors cause the vehicle to vibrate irregularly, thereby causing the on-board camera to shake irregularly, which in turn affects image collection; at the same time, the dust concentration underground is relatively large, which affects the visibility underground, thereby affecting the clarity of the obstacle recognition image collected by the camera; in addition, due to the deviation of the camera's focus on obstacle recognition, the obstacle recognition image taken by the camera is relatively blurred, affecting the accuracy of obstacle recognition; the above three factors will all lead to a decrease in the accuracy of obstacle recognition in front of the vehicle in the underground lane of the coal mine. In order to comprehensively evaluate the influence of the above three factors on obstacle detection, the present invention provides the following specific steps, including:

[0044] Step 1: collecting vehicle vibration parameters of a vehicle traveling in a mine, wherein the vehicle vibration parameters include the vehicle vertical acceleration and lateral translation acceleration;

[0045] The vehicle vertical acceleration A is collected by a three-axis acceleration sensor installed on the moving vehicle. Y and lateral translation acceleration A X , A XA is the acceleration value of the vehicle perpendicular to the direction of travel of the vehicle and parallel to the horizontal plane, Y It is the vehicle acceleration value in the vertical direction of the moving vehicle.

[0046] Step 2: Perform correlation analysis on the vehicle vibration parameters to generate the vehicle vibration speed variation coefficient;

[0047] The vertical acceleration of the vehicle is A Y and lateral translation acceleration A X Correlation analysis is performed to generate the vehicle vibration speed variation coefficient CDX, based on the formula: Among them, ε is the vehicle acceleration jitter factor, and the value range of ε is [0.35,3.88]. The vehicle jitter speed variation coefficient reflects the severity of the lateral and longitudinal comprehensive jitter of the moving vehicle. The vehicle jitter speed variation coefficient is positively correlated with the severity of the comprehensive vehicle jitter. The larger the vehicle jitter speed variation coefficient CDX is, the more severe the comprehensive vehicle jitter is, and the higher the correlation with the obstacle image collected by the on-board camera is.

[0048] Step 3: collecting a reference image of the traveling vehicle as a baseline reference, collecting a comparison image of the traveling vehicle traveling in a mine as a comparison reference, and collecting pixel values ​​of reference objects in the reference image and the comparison image;

[0049] A camera and a reference object are set outside the moving vehicle. The distance between the reference object and the camera is greater than 50 cm and less than 150 cm. When the PM2.5 mass concentration is less than 50 μg·m 3 When the relative humidity of the air is less than 60%, that is, when the visibility is high, the camera is used to shoot a reference object to generate a reference image. The reference object in the reference image has M rows and N columns of pixels, and the grayscale value of the pixel in the m0th row and n0th column of the reference object in the reference image is recorded as I (m0, n0); during obstacle image recognition and detection, the camera is used to shoot a reference object to generate a comparison image. The reference object in the comparison image also has M rows and N columns of pixels, and the grayscale value of the pixel in the m0th row and n0th column of the comparison image is recorded as G (m0, n0). The grayscale value calculation formula of the present invention is: Gray = 0.114B + 0.587G + 0.299R, wherein Gray represents the grayscale value, B, G, and R are the blue, green, and red color values ​​of the pixel, respectively, and the value range is [0, 255].

[0050] Step 4: Perform correlation analysis on the grayscale values ​​of the reference objects in the reference image and the comparison image to generate a reference environmental impact coefficient;

[0051] The correlation analysis is performed on the grayscale values ​​of the reference objects in the reference image and the comparison image to generate the reference environmental impact coefficient HYX, based on the following formula: The reference environmental impact coefficient reflects the influence of the underground mine environment on the shooting clarity. The reference environmental impact coefficient is positively correlated with the obstacle image clarity, that is, the larger the reference environmental impact coefficient, the clearer the obstacle image.

[0052] Step 5: Collect obstacle recognition images taken by the camera of the vehicle traveling in the mine;

[0053] The obstacle recognition image has X rows and Y columns of pixels, and the brightness of the pixel in the i-th row and the j-th column is represented by f(i, j).

[0054] Step 6: Perform correlation analysis on the obstacle recognition image to generate pixel mean, pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation. Perform correlation analysis on the pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation to generate a no-reference image clarity evaluation coefficient.

[0055] Perform correlation analysis on the obstacle recognition image to generate pixel mean L, pixel grayscale distribution discrete value H, horizontal pixel change degree T X , vertical pixel change degree T Y , the formula based on is:

[0056]

[0057] The pixel mean reflects the average brightness of the obstacle recognition image, the pixel grayscale distribution discrete value H reflects the brightness discreteness of the pixel points of the obstacle recognition image, and the horizontal pixel variation T X It reflects the brightness change degree of pixels separated by two pixels in the horizontal direction of the obstacle recognition image. The vertical pixel change degree T Y It reflects the degree of brightness change of pixels separated by two pixels in the vertical direction of the obstacle recognition image.

[0058] The correlation analysis of the discrete value of pixel grayscale distribution, horizontal pixel variation, and vertical pixel variation is performed to generate the reference-free image clarity evaluation coefficient TQP, based on the following formula: Among them, α is the brightness discrete influence factor of the obstacle recognition image, β is the gradient discrete influence factor of the obstacle recognition image, and β>α≥6.22. The non-reference image clarity evaluation coefficient TQP reflects the clarity of the obstacle recognition image. The larger the non-reference image clarity evaluation coefficient TQP is, the higher the clarity of the obstacle recognition image is.

[0059] Step 7: Conduct correlation analysis on the clarity assessment coefficient of the non-reference image, the reference environmental impact coefficient, and the vehicle vibration speed variation coefficient to generate a comprehensive clarity assessment index for mine vehicle image recognition. Compare the comprehensive clarity assessment index for mine vehicle image recognition with the impact threshold and output the accuracy level of obstacle image recognition in front of vehicles traveling in coal mine tunnels.

[0060] The correlation analysis of the no-reference image clarity evaluation coefficient, the reference environmental impact coefficient, and the vehicle vibration speed change coefficient is performed to generate the mine vehicle image recognition comprehensive clarity evaluation index KSQ, based on the following formula: The comprehensive clarity evaluation index of mine vehicle image recognition reflects the accuracy of obstacle image recognition in front of vehicles traveling in underground coal mine tunnels.

[0061] The impact threshold is set to θ. When KSQ>θ, the accuracy level of the output obstacle image recognition of the vehicle in front of the coal mine tunnel is level 1, and the accuracy of the output obstacle image recognition is greater than 95%; when When the accuracy level of the output obstacle image recognition of the vehicle in front of the coal mine tunnel is level 2, the accuracy of the output obstacle image recognition is less than or equal to 95% and greater than 80%. When the accuracy level of the output obstacle image recognition in front of the vehicle traveling in the underground coal mine tunnel is unqualified, the accuracy of the output obstacle image recognition is less than or equal to 80%, and it is not recommended to be adopted.

[0062] See also Figure 2 The present invention also provides a system for detecting obstacles in front of vehicles traveling in underground mine lanes, which is used to execute a method for detecting obstacles in front of vehicles traveling in underground mine lanes, comprising:

[0063] An acceleration acquisition module, which is used to collect vehicle vibration parameters of vehicles traveling in the mine;

[0064] A first analysis module, the first analysis module is used to perform correlation analysis on vehicle vibration parameters to generate a vehicle vibration speed variation coefficient;

[0065] An environment change model establishment module, the environment change model establishment module is used to set a camera and a reference object outside the moving vehicle;

[0066] An environment change acquisition module, the environment change acquisition module is used to acquire a reference image of a moving vehicle as a baseline reference, and to acquire a comparison image of the moving vehicle when it is traveling in a mine as a comparison reference;

[0067] A second analysis module, the second analysis module is used to perform correlation analysis on the grayscale values ​​of the reference image and the comparison image to generate a reference environmental impact coefficient;

[0068] An image recognition acquisition module, which is used to acquire obstacle recognition images taken when the vehicle is in motion;

[0069] a third analysis module, the third analysis module being used to perform correlation analysis on the obstacle recognition image, generate pixel mean, pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation, and perform correlation analysis on the pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation to generate a no-reference image clarity assessment coefficient;

[0070] A comprehensive analysis module, which is used to perform correlation analysis on the vehicle jitter speed variation coefficient, the reference environment impact coefficient, and the non-reference image clarity evaluation coefficient output by the first analysis module, the second analysis module, and the third analysis module, and generate a comprehensive clarity evaluation index for mine vehicle image recognition;

[0071] A comparison module, wherein the comparison module is used to compare the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine with the impact threshold, and output the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine; when the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine is greater than the impact threshold, the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine is output as level one; when the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine is less than the impact threshold, the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine is output as level two.

[0072] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0073] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0074] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0075] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A method for detecting obstacles in front of a vehicle traveling in a coal mine tunnel, which is suitable for collecting images and identifying obstacles through a vehicle-mounted camera, and is characterized in that: The specific steps include: Step 1: collecting vehicle vibration parameters of a vehicle traveling in a mine, wherein the vehicle vibration parameters include the vehicle vertical acceleration and lateral translation acceleration; Step 2: Perform correlation analysis on the vehicle vibration parameters to generate the vehicle vibration speed variation coefficient; Step 3: collecting a reference image of the traveling vehicle as a baseline reference, collecting a comparison image of the traveling vehicle traveling in a mine as a comparison reference, and collecting pixel values ​​of reference objects in the reference image and the comparison image; Step 4: Perform correlation analysis on the grayscale values ​​of the reference objects in the reference image and the comparison image to generate a reference environmental impact coefficient; Step 5: Collect obstacle recognition images taken by the camera of the vehicle traveling in the mine; Step 6: Perform correlation analysis on the obstacle recognition image to generate pixel mean, pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation. Perform correlation analysis on the pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation to generate a no-reference image clarity evaluation coefficient. Step 7: Conduct correlation analysis on the clarity evaluation coefficient of the non-reference image, the reference environmental impact coefficient, and the vehicle jitter speed variation coefficient to generate a comprehensive clarity evaluation index for mine vehicle image recognition, and compare the comprehensive clarity evaluation index for mine vehicle image recognition with the impact threshold to output the accuracy level of the image recognition of obstacles in front of vehicles traveling in coal mine tunnels; In step 1, the vertical acceleration A of the vehicle is collected by a three-axis acceleration sensor installed on the moving vehicle. Y and lateral translation acceleration A X , A X A is the acceleration value of the vehicle perpendicular to the direction of travel of the vehicle and parallel to the horizontal plane, Y is the vehicle acceleration value in the vertical direction of the moving vehicle; The vertical acceleration of the vehicle is A Y and lateral translation acceleration A X Correlation analysis is performed to generate the vehicle vibration speed variation coefficient CDX, based on the formula: Where ε is the vehicle acceleration jitter factor, and the value range of ε is [0.35, 3.88]. The vehicle jitter speed variation coefficient reflects the severity of the lateral and longitudinal comprehensive jitter of the moving vehicle; In step 3, a camera and a reference object are set outside the moving vehicle, and the distance between the reference object and the camera is greater than 50 cm and less than 150 cm. When the mass concentration of PM2.5 is less than 50 μg·m 3 When the relative humidity of the air is less than 60%, the reference object is photographed by the camera to generate a reference image, the reference object in the reference image has M rows and N columns of pixels, and the grayscale value of the pixel in the m0th row and n0th column of the reference object in the reference image is recorded as I(m0,n0); during obstacle image recognition and detection, the reference object is photographed by the camera to generate a comparison image, the reference object in the comparison image also has M rows and N columns of pixels, and the grayscale value of the pixel in the m0th row and n0th column of the comparison image is recorded as G(m0,n0); The correlation analysis is performed on the grayscale values ​​of the reference objects in the reference image and the comparison image to generate the reference environmental impact coefficient HYX, based on the following formula: The reference environmental impact coefficient reflects the influence of the underground mine environment on the shooting clarity; In step 5, the obstacle recognition image has X rows and Y columns of pixels, and the brightness of the pixel in the i-th row and j-th column is represented by f(i, j); Perform correlation analysis on the obstacle recognition image to generate pixel mean L, pixel grayscale distribution discrete value H, horizontal pixel change degree T X , vertical pixel change degree T Y , the formula based on is: The pixel mean L reflects the average brightness of the obstacle recognition image, the pixel grayscale distribution discrete value H reflects the brightness discreteness of the pixel points of the obstacle recognition image, and the horizontal pixel variation T X It reflects the brightness change degree of pixels separated by two pixels in the horizontal direction of the obstacle recognition image. The vertical pixel change degree T Y It reflects the brightness change of pixels separated by two pixels in the vertical direction of the obstacle recognition image; The correlation analysis of the discrete value of pixel grayscale distribution, horizontal pixel variation, and vertical pixel variation is performed to generate the reference-free image clarity evaluation coefficient TQP, based on the following formula: Wherein, α is the brightness discrete influence factor of the obstacle recognition image, β is the gradient discrete influence factor of the obstacle recognition image, and β>α≥6.22, and the no-reference image clarity evaluation coefficient TQP reflects the clarity of the obstacle recognition image; The correlation analysis of the no-reference image clarity evaluation coefficient, the reference environmental impact coefficient, and the vehicle vibration speed change coefficient is performed to generate the mine vehicle image recognition comprehensive clarity evaluation index KSQ, based on the following formula: The comprehensive clarity evaluation index of mine vehicle image recognition reflects the accuracy of obstacle image recognition in front of vehicles traveling in underground coal mine tunnels.

2. A system for detecting obstacles in front of vehicles traveling in underground mine lanes, used to execute the method for detecting obstacles in front of vehicles traveling in underground mine lanes as claimed in claim 1, characterized in that: include: An acceleration acquisition module, which is used to collect vehicle vibration parameters of vehicles traveling in the mine; A first analysis module, the first analysis module is used to perform correlation analysis on vehicle vibration parameters to generate a vehicle vibration speed variation coefficient; An environment change model establishment module, the environment change model establishment module is used to set a camera and a reference object outside the moving vehicle; An environment change acquisition module, the environment change acquisition module is used to acquire a reference image of a moving vehicle as a baseline reference, and to acquire a comparison image of the moving vehicle when it is traveling in a mine as a comparison reference; A second analysis module, the second analysis module is used to perform correlation analysis on the grayscale values ​​of the reference image and the comparison image to generate a reference environmental impact coefficient; An image recognition acquisition module, which is used to acquire obstacle recognition images taken when the vehicle is in motion; a third analysis module, the third analysis module being used to perform correlation analysis on the obstacle recognition image, generate pixel mean, pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation, and perform correlation analysis on the pixel grayscale distribution discrete value, horizontal pixel variation, and vertical pixel variation to generate a no-reference image clarity assessment coefficient; A comprehensive analysis module, which is used to perform correlation analysis on the vehicle jitter speed variation coefficient, the reference environment impact coefficient, and the non-reference image clarity evaluation coefficient output by the first analysis module, the second analysis module, and the third analysis module, and generate a comprehensive clarity evaluation index for mine vehicle image recognition; A comparison module, wherein the comparison module is used to compare the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine with the impact threshold, and output the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine; when the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine is greater than the impact threshold, the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine is output as level one; when the comprehensive clarity evaluation index of the image recognition of vehicles traveling in the mine is less than the impact threshold, the accuracy level of the image recognition of obstacles in front of vehicles traveling in the underground lanes of the coal mine is output as level two.

Images